General purpose

I’m creating Rmds for each of the datastreams used in this project. Because our deployments were a little complex, and both SWAP and Firesting datasets were somewhat cobbled together, it’s important to document as many details as possible as clearly as possible in order to streamline QC, clearly ID and explain any data decisions, and make the methods section easy to write.

Sapflow deployment

Sapflow sensors were deployed well prior to any flooding in three species of trees across TEMPEST plots. I honestly don’t know much about their deployment, which is okay because these data, in contrast to TEROS/Firesting/SWAP datasets is being pulled in as L1 curated data from GDrive meaning initial QC has already taken place (no need to edit timestamps, etc).

Raw data

Let’s first take a look at all the raw data.

QC

QC Step 1: Remove deep sensors

We will scrub the sensors with a “D” suffix which indicates they are deep sensors. Since these are primarily used for calculating areal rates which we aren’t doing, we’ll keep things apples to apples and remove them.

QC Step 2: examine voltages

Based on sapflow voltages, Control and Estuarine plots seem good to go, but there is a power issue in Freshwater for all loggers prior to 6/4 that seems to get fixed right before the event. Let’s use the first datetime of good voltages in Freshwater as the start-date for our sapflow, which works out well since the redox sensors were installed after this date. The time when voltages resets is 2023-06-04 12:00:00. There’s also some sort of step change in the Estuarine plot, which fortuitously happens around midnight on 6/13. Since we really only care about running sapflow out to 6/13 to overlap with veg measurements, let’s cut that too:

We are getting closer, but there are some days tht re visually apparent anamolies that we should scrub. These include June 12, and then anything after ~ June 21

After removing those issues as well, this is our cleaned dataset after Step 2

QC Step 3: visual examination

These data are starting to look good, so let’s look at individual sensors and see if there are any potential issues with jumps or trends or any other non-normalities we should pay attention to. Based on visual identification, the following sensors are sus for the following reasons:

  • F10: non-linear behavior
  • F20: high (values above 0.7)
  • C10: high (values above 0.8)
  • S2: linear trend
  • S8: linear trend

Final data are exported as 240717_sapflow_final_unbinned.csv in the data folder